Cities and Sustainability: Is Intensification Good Policy?

Evidence on urbanisation and conservation from New South Wales ...

Squeezing up to save the environment
This post examines the idea that we can promote sustainability by increasing the densities of large cities around their centres.  This compact city paradigm presumes that we can reshape the consumption of citizens in environmentally benign ways by reshaping the cities they live in. 

The sustainability challenge is the challenge of consumption: how much and what we consume drives our impact on the planet.  But presuming that by enforcing urban intensification we will transform ingrained patterns of consumption in favour of the environment may be a step too far.  Will obliging more citizens to live at higher densities in smaller dwellings around city centres really pave the way to environmental salvation?

Some evidence of urban impacts

The Australian Conservation Foundation is committed to ecological sustainability, tackling the social and economic causes of environmental problems.  Among other things, the Foundation publishes the online Australian Consumption Atlas. This is a useful source for addressing the role of urbanisation and urban form.

The Atlas is based on methodology which traces the direct and indirect demands on the environment of different goods and services.  Consumption patterns from Household Expenditure Surveys are related to household size and type, members’ age structure, incomes and education, and the statistical areas they live in. Using this information the environmental impacts of individuals living in different areas can be mapped. 

Three indicators of impact are displayed in the atlas: tonnes of greenhouse gas emitted, litres of water consumed, and ecological footprint.  The latter estimates the area of resources required to support a person’s lifestyle.  You can read more about the methodology here.

The data underlying the atlas is dated – based on the 2001 Census and 1999 Household Expenditure Survey, among other things.  But I do not expect the relativities it demonstrates, or the conclusions it supports, to have changed much.

Cities don’t consume; people do

Here is the authors' key conclusion. Our urban planners, designers, and politicians should consider carefully:

despite the lower environmental impacts associated with less car use, inner city households outstrip the rest of Australia in every other category of consumption. Even in the area of housing, the opportunities for relatively efficient, compact living appear to be overwhelmed by the energy and water demands of modern urban living, such as air conditioning, spa baths, down lighting and luxury electronics and appliances, as well as by a higher proportion of individuals living alone or in small households.

In each state and territory, the centre of the capital city is the area with the highest environmental impacts, followed by the inner suburban areas. Rural and regional areas tend to have noticeably lower levels of consumption.

(Consuming Australia: Main Findings, 2007, Australian Conservation Foundation, p.10)

Looking inside Sydney

I explored the indicators for different parts of Sydney.  Here are some results.



Indicators of Environmental Impacts: Sydney Centre and Suburbs

People in Inner Sydney generate 92% more greenhouse gas than the New South Wales Average, and well over twice as much as people in the lower income western suburbs, like Penrith and Blacktown. The levels are a bit higher for people in the more prosperous northern suburbs. Despite proximity to major employment centres, and an efficient commuter rail service, the consumption patterns of Willoughby and Ku-ring-gai residents generate high levels of air pollution.


Looking East to Sydney CBD
(Source: www:freeaussiestock.com)

A similar pattern is evident for water consumption – residents of the hot, dry, western suburbs account for the least consumption, Inner and North Sydney residents the most.  They also have the biggest ecological footprint.

 

So what does this tell us?
The lesson is not necessarily that location in the CBD is less sustainable; but that the lifestyle associated with it is.

I have discussed the potential inefficiency of small, multi-unit dwellings elsewhere.  Over and above that, the high cost of redevelopment in central locations calls for housing construction strategies that add little to sustainability. 

One strategy is to build to modest standards.  This keeps the price down and rental yield up for investors; or creates opportunities for ownership by low income earners.  Another strategy is to adopt high standards of fit-out and install luxury appliances in favoured locations to make multi-unit dwellings attractive to wealthier households. 

Neither option is particularly environmentally sympathetic. 

Smaller is still better

I also reviewed the indicators for smaller cities and towns in New South Wales.  (In some cases these included surrounding rural settlement). 


Indicators of Environmental Impacts: New South Wales Towns and Small Cities

This suggests that smaller towns hold the key to environmentally sustainable lifestyles, even more than city suburbs.  For example,  Coffs Harbour's 73,000 residents generate greenhouse emissions at 88% of the state average, and just 46% of inner Sydney residents.  They consume water at 81% of the State rate (and 60% of North Sydney), and have an ecological footprint just 60% of their inner Sydney counterparts.  Similar patterns are evident in coastal settlements like Byron Bay (33,000 residents), Ballina (42,000), and Port Macquarie (77,000) and inland towns such as Griffith (26,000), Tamworth (60,000), and Wagga Wagga (64,000).

What does it all add up to?

A simple overview can be derived by summing the percentage deviations of each area from the New South Wales average across the three measures. Admittedly this is a course approach: it weights each indicator equally, and ignores differences in how much centres vary across each individually.  Nevertheless, it provides a sufficiently meaningful overview to confirm that towns and small cities are generally more sustainable than a large city, and that the suburbs perform better than the inner city.


Summary Index of the Environmental Impact of Urbanisation
 

Explaining the sustainability dividend of small towns
There can be any number of explanations for this, the obvious one being that it is all about income.  Perhaps the advantages of lifestyles outside Sydney simply reflect lower average incomes in smaller cities and towns.  As people become more affluent or seek more income, they migrate into the main cities taking their high consumption expectations with them; or by living in large cities they are more likely to earn - and consume - more.

Conversely, living in smaller cities and settlements may reflect lifestyle preferences which are intrinsically less environmentally intrusive.  At the same time. small settlements make less travel demands given the greater proximity to work, shopping, service, and recreation opportunities.  In addition, lower density housing may provide more opportunities for passive energy efficiency, directly reducing resource consumption for comparable activities.  

Flawed policy

Until we know more, however, we need to avoid the trap of determinism.  It would be short-sighted simply to invert the current paradigm, for example, and decide that policies to encourage people to live outside large cities and city centres will somehow enhance sustainability.

Ultimately, how we live is more important than where we live.  What the evidence here confirms, though, is that under current patterns of consumption promoting large scale urban consolidation is flawed as environmental as well as urban policy. 

The Answer Is Urban Consolidation – What Was The Question?

Perpetuating the Myth

The Green Party is perpetuating the claim that development beyond Auckland’s “city limits” imposes a high cost on ratepayers.  A spokesperson claims that the current Auckland plan which allows for some new development outside the current urban area, “will cost ratepayers $42b billion to 2042, an annual levy of $200 per ratepayer”  according to a report in the New Zealand Herald.   

But is just so happens that  study on which these calculations are based is a flawed commissioned report  rather than a peer reviewed academic study (Roman Trubka, Peter Newman and Darren Bilsborough (2008) Assessing the Costs of Alternative Development Paths in Australian Cities, Curtin University Sustainability Policy Institute, Fremantle, Report commissioned by Parsons Brinckerhoff Australia). 

Oops – Contradictory Claims

The authors of the Curtin report acknowledged at the outset that

"The challenge ...  is that infrastructure costs are so heavily dependent on area-specific values.  For instance, road costs among different prospective development areas may vary based on the necessity for major arterial roads, costs for sewerage and water infrastructure could vary immensely depending on terrain and trenching conditions, and many infrastructure components will differ depending on the level and degree of excess capacity” (p.4)

So why did they try to develop a generic tool for estimating the cost of urban development in Australian cities based on a mishmash of evidence from different cities and suburbs in Australia and the United States?  And why would anyone even contemplate applying such “findings” to Auckland with its distinctive physical geography, so different from its Australian counterparts? 

A Quick Critique

The Productivity Commission actually considered the study, among others, in a brief review of housing costs and urban form (Appendix B of the final report).  It noted substantive differences in the physical and social settings  behind the data assembled to support the  study’s claim to some sort of universal cost relationship between development and distance from the city centre.

And there are glaring methodological deficiencies:

“An obvious one is mixing discount rates (zero for infrastructure capital costs, 7% for transport-related costs, and 3% for health and emission costs), and omitting operating costs for some items (non-transport infrastructure) and not others (pp. 295-296)

To these flaws can be added the assumption of a cost of Aus$170/tonne for carbon emissions when the carbon floor price set by the Australian government (of $15) has since been rescinded and figures at or below $10.00 may be more appropriate based on today’s European prices.  So the environmental argument is seriously overstated.

And the analysis fails to deal with the costs of expanding the capacity of ageing infrastructure in long-established urban areas, of remediating services designed for far lower loadings than they are now expected to sustain, of the health impacts of apartment living in an increasingly brown – not green – environment, and of reductions in the physical and socialresilience of high density and often congested urban areas in the face of possible natural disasters or infrastructure failures.

Penalising the Household - is that Socially Sustainable, or Politically Justified? 

Even if it can be proven that the balance of public benefits favours medium or high density living, is there any evidence that such savings will not be offset by the better affordability of traditional suburban housing and the benefits residents derive from living into it?

Putting aside  flawed data and methodology for the moment, the results indicate that 70% of the differences in costs between decentralised and central locations is attributable to travel and transport.  Over half of these comprise travel costs and  time carried  by households.  If we take these private costs out of the equation the authors' estimate of the difference  between centralised and decentralised development falls by 40%.  

The resulting "present cost" for the average household (whatever that might be) of A$22,000 is easily  justified by savings on land and housing in “outer” areas, the benefits households get  from  additional space, greater choice over housing style, and the security and community benefits of suburban environments.

So who pays if we deny people the choice of living in medium to low density housing?  Mainly new households through exclusion from household ownership, or commitment to punitive mortgages, or through the insidious extension of housing poverty through ever higher income brackets. 

So what about the Auckland case: where does the evidence really lie?

Surprisingly-- given the obstinacy of the planners and politicians pushing the consolidation barrow --   no-one has actually done the analysis required to determine the relative economic benefits of different urban development paths for Auckland.  

A technical analysis of the gaps in the Auckland Regional Growth Strategy made the point that the planning  model that informed it was hardly up to the task.  The principal conclusion that came from using the Regional Council's land use and transport  model was that there is “little [identified] economic difference between growth options”.(McDermott Fairgray Ltd (1999) Gap Analysis, Review and Recommendations: Auckland Regional Growth Strategy, Technical Report, Auckland Regional Growth Forum )

The failure of the model to demonstrate economic differences between alternative urban forms was used to suggest that intensification imposes no additional costs than traditional  decentralised development.  Of course, the converse is true – although it has been conveniently ignored: there were no demonstrable economic benefits from consolidation or net cost penalties to decentralisation.  This suggests that it would make most sense to let the market prevail, subject  to broad environmental standards and fiscal constraints.   

The  conclusion  that consolidation was the best option for Auckland ignores other shortcomings  in the  model that could  tip the balance  in favour of strategic decentralisation:

• The failure to actually define realistic alternatives that would  clearly demonstrate economic differences;
• A failure to evaluate the marginal rather than average impacts of differences in urban form;
• The failure to identify the costs of implementation.
• Ambiguous measurement (both omissions and double counting);

To this list we can add underestimation of the high infrastructure and development costs associated with brownfield development and urban consolidation.  These are turning up today in high financial and development contributions for inner city projects.

Calling for  Consolidation – a Case of Artificial Intelligence

So why is the Auckland Spatial Plan so fixated on consolidation –despite the begrudging lip service the final version pays to decentralisation (a small concession to market reality that appears to have  upset  the Green spokesperson)?

I can only think it is "artificial intelligence": if enough people say the same thing, it must be right.  Consensus becomes an excuse for lack of evidence, critical analysis, or even common sense.  Groupthink prevails,: a phenomenon defined by psychologist Irving Janis as:

A mode of thinking that people engage in when they are deeply involved in a cohesive in-group, when the members' strivings for unanimity override their motivation to realistically appraise alternative courses of action (Janis, I L (1972). Victims of Groupthink Houghton Mifflin p. 9)

Contrary evidence is dismissed while reports favouring an emerging consensus, such as the Curtin one, obtain a degree of currency which, while unjustified,  plays into the hands of policy makers looking for easy (or ideologically comfortable) answers to difficult problems.

And so we blunder on, potentially building our cities on myth and misconception and reinforcing the gap betwen generations as we do it.

---

Being dense about dwellings: check the numbers!

Good news, bad news
In a previous posting I suggested that in New Zealand we are heading into the perfect housing storm. 

Now we have news that house prices and rentals are on the climb again, although stocks remain tight, as an annual inflation rate of 5.3% hits a 21 year high.  The economists are suggesting this is good news, although it means interest rates may have to be pushed up sooner than expected.

Well the bad news is that the housing crisis might just have worsened. 

Sure, its not an across-the-board crisis, but it is very real to large and important sections of our population.  Lack of housing affordability remains a threat to social sustainability and economic recovery.  So how are we responding to the threat -- or perhaps now the reality -- of a perfect housing storm?  What provisions are we making in our urban plans?

Smaller boxes – bigger footprint
Urban planners are still more preoccupied with fitting more dwellings into smaller areas than they are with responding to people's needs for housing.  It might help shift this fixation to point out that the preferred compact city solution is not only socially destructive, because it doesn't reflect need and does nothing for affordability, but it is also environmentally short-sighted.

Think about the metrics.

Take 100 people and house them at 1.5 residents per dwelling.  That’s arbitrary, but it reflects a widespread expectation that most new dwellings will house smaller households in central locations. 

In the interests of sustainability, let’s assume the resulting 67 dwellings are small, so that we can fit more of them onto less land.  Say, 120 sq meters per dwelling.  That totals 8,000 sq metres or thereabouts (more if we count the common areas in apartment buildings), 80 sq metres per person.  It’s also 67 kitchens, 67 lounges, maybe 67 media centres, at least 67 bathrooms, maybe some additional lighting for common areas and even some lifts.

Now take 100 people and fit them in at 3 people per dwelling, terraces, duplexes or fully detached houses.  Let’s make the dwellings bigger, say 200 sq metres.  We now need only 33 dwellings, 6,600 sq metres of dwelling, or 66 sq metres per person.  Less space per person, sure, but that's okay because now we need just half the kitchens, bathrooms, lounges and media centres.  However we look at it, we’ve used a lot less resources and have a spare 1,400 sq metres for open space, extra gardens, courtyards, whatever.  And with the capacity for extra bedrooms, we have much more flexible housing stock.

So which is the more sustainable?  Surely bigger dwellings with higher occupancies.  Surprised?

Can we plan for higher occupancies?
Now, we can’t engineer household size, can we?  Well, actually we already do.  With a housing shortfall we now require young adults to stay longer with their parents, force singles to move in with others,  require couples to take on boarders, or even promote multi-family living, all boosting occupancies.

So let’s at least understand that building more, smaller dwellings, especially medium- or high-rise apartments, does not necessarily deliver sustainable urban settlement, nor does it provide the flexibility to make the higher occupancy "solutions" we force on people easy to live with.

Larger dwellings do allow for diverse living arrangements, though; more multi-generational living, more non-family households, more sharing.  Like them or not, such arrangements are likely to increase, if only in response to the affordability issues we seem intent on entrenching.

So what’s happening to demand?
So why are planners trying to put more people into smaller dwellings anyway?  How relevant is the expectation that average household size will be smaller in the future than it has been in the past?

Most forecasts of housing “demand” simply extrapolate diminishing occupancy across demographic projections.  its all about the coefficients, and the assumption that household structures won’t change much in the medium to long-term. 

Well, it’s not that simple.

Things like an unexpected boom in the dissolution of relationships over the past three or four decades, the rapid growth in migration, and the recent stabilisation and even reversal in occupancy rates undermine the conceit that we can accurately forecast the structure, preferences, and behaviour of households 20 or 30 years hence.  If that's the case, why are our prescriptions for housing increasingly rigid?

Projecting household types
To understand this let’s stay with the current ”best”  projections of what households might look like in the future, and think about the implications for housing.

Statistics New Zealand (SNZ) medium projections to 2031 indicate that families with children will account for a minority of household growth in our main cities (see chart).  The figures may even shrink in Wellington and Christchurch.  According to this projection, they will make up 28% of new households in Auckland, though, so we could still need over 71,000 new dwellings for families there.  It’s reasonable to expect that detached housing will still work best for them.

Household Category Projections, Statistics New Zealand

Couples will account for more growth, though, maybe 36% of new households in Auckland according to SNZ, and singles for 32%.  So let's think about the preferences of the small household segment. 

So what will the small household segment look  like?
To get a feel for this, I divided the SNZ age projections into four (setting aside the main family age cohorts) : young adults (aged 20-29), empty nesters (the kids have left home, aged 50-64); early retirees (65-79), and later retirees (80+).  These are the groups most small households will come from.  But they have quite different housing preferences, so the nature of future demand for smaller dwellings depends on which ones grow the most.

Age-Based Housing Demand Segments (based on SNZ Projections)

So who will dominate growth?
Empty nesters and retirees will dominate the demand for new houses.  And these are not usually people who want to move into small, centralised apartments, at least not as a primary residence. 

Many of them have significant financial equity in their existing homes and emotional equity in their neighbourhoods.  If they move into smaller dwellings, they won’t be that small!  They will expect them to be well appointed and well located, probably close to where they already live. 

They won't want high or even medium rise.  And they are  likely to seek three or four bedrooms.  They will need the space to maintain active  lives into their seventies and eighties, more so than past generations.  They will be accommodating visiting family and friends; they will need offices, hobby areas, workshops, and storage. 

Here’s a model to take seriously if we are serious about sustainability

And as the baby boomers eventually become less independent, we might expect them to head into retirement villages, already a booming – and highly sustainable – form of housing.

In fact, we should look seriously at retirement villages if we want to understand the sorts of arrangements that could dominate new housing demand over the next 30 years.  Here, the market seems to have got it right. 

They offer varied living arrangements – detached and semi detached housing, terraces, apartments, and even on-site nursing facilities.  They offer medium density living with plenty of green space and gardens; common areas and shared facilities for recreation and leisure; plenty of on-site activity to cut down transport needs but also on-site parking to reflect the realities of modern living.  They achieve density and sustainability with style.  And – there must be a lesson here – they do it overwhelmingly in suburban if not city edge localities. 

So let's not assume that rising house prices mean a return to business as usual.  Far from it - freeing up the housing market must remain a top priority if the economy is in recovery mode.  And let's start looking to the suburbs  –and beyond  – for the housing solutions that might just help it stay that way.

Cities in search of resilience

Resilience: the ability to recover quickly from setbacks (Encarta)

An age of extreme events?

Without debating whether an increase in the frequency of extreme events reflects climate warming , such events can be catastrophic when they impact on densely populated areas.  Natural disturbances, whether geophysical (tsunami, earthquakes, mudslides) or climatic (flooding, hurricane strength winds, tidal surges), become disasters if they strike heavily populated centres. 

So do human acts of aggression.  The tactic of terrorising civilian populations taken to new heights in the bombing raids of the Second World War and adopted by today’s extremists is most effective – and destructive - when directed at the heart of major cities.

 

 

Promoting preparedness

So how do we respond, especially given the expectation that we face an increase in such events?

We can prepare ourselves individually by sensible precautions. House design, construction, maintenance can help in high risk areas.  Having household plans and resources for escape, survival, and recovery is becoming more common.  As communities we can build our collective emergency response and recovery capacity.  We can also look to our hinterlands to ensure that land use practices -- clearances, monocultures, river straightening, irrigation, and dams -- do not precipitate major events such as dust storms or floods that impact on cities downwind and downstream.

The Civil Defence Emergency Management Act 2002 increased the New Zealand government’s focus on risk reduction, hazard avoidance, and community readiness.  In particular, Part 1 (3) (d) requires:

local authorities to co-ordinate, through regional groups, planning, programmes, and activities related to civil defence emergency management across the areas of reduction, readiness, response, and recovery, and encourage co-operation and joint action within those regional groups;

The effects of climate change, presumably including the potential for violent storms and inundation, are matters to which people exercising powers under the Resource Management Act should have particular regard (Part 2 7 (i)).

Vulnerable by design

There is an imperative in legislation in New Zealand, then, for local and regional councils to consider hazard mitigation and risk avoidance in our urban planning and design.  This is so internationally.

It is not coincidence that many – if not most – major cities in the world are built on rivers or at the coast given their origins as nodal points, or  on fertile flood plains, in the lee of mountains, between mountains and sea, even on fault lines.  They are consequently built across unstable and vulnerable sites in many instances.  This is should be a fundamental consideration in our urban design, architecture, and engineering. [1]

So here are some reasons why we might seriously question the compact city paradigm which so influences planning and urban design today:

1. It relies on sophisticated, centralised interdependent systems of services.  This creates greater capacity for disruption when any one part fails.  Economies of scale in utilities may come with increased risk of failure under duress.  This applies to sewage treatment infrastructure, communications, water, energy distribution, and power supplies.  It also applies to public transport systems.
2. Poorly designed intensification reduces permeable surfaces, intensifying flood impacts.
3. Converting brownfield and even greenfield sites (such as undeveloped urban space) to housing or mixed use reduces the safety valve of open space and increases vulnerability associated with the concentration of buildings and populations.
4. Crowding more people into smaller spaces around constrained road capacity reduces prospects for rapid evacuation from the city or into safe structures and areas.
5. Lifting the density of buildings increases the consequential impacts of severe events by such things as the collapse of structures, the spread of fire, and the transmission of disease.
6. Mixing uses increases the risk of injury and destruction when people live close to premises where hazardous and flammable goods may be stored.  Gas, chemical cleaners, and fuel are obvious examples.
7. Reducing the space available reduces the capacity of people – households and communities – to fend for themselves, particularly if the consequences of a disruptive event are prolonged.

Looking over these issues, it is unsurprising that our past history of increasing prosperity was a history of reducing urban densities even as rural-urban migration pushed up city populations. What is surprising is that we seem to have given up the quest to make this same movement -- essentially a de facto public health programme -- work in a resource constrained environment.

Back to the future?

One of the drivers of early town and country planning was the desire to protect public health, with zones separating industry from where people lived.  A healthy workforce was a productive workforce, so it made sense to reduce the exposure of people to industrial pollution.  There were also public health benefits from getting families out of high density slums into something approximating a rural lifestyle with access to space, gardens, and parks.  The resulting residential areas – the suburbs -- came to be highly valued in the 20^th century.  Many people still value them, even in a post-industrial age.

But in today’s quest to preserve city edges, to support public transit, and curtail car use planners have moved to reinstate higher urban densities around existing city centres, denigrating suburban life as “sprawl”, and downplaying the new risks that revisiting the old ways raise.

Planning for resilience

We may have to rethink these revisionary ideals in the face of reality.  A better understanding of resource constraints and the need for diversity may mean that we shouldn't look at expanding our cities in the uncritical way we did in the past, providing large plots for small households.  But for many people, and perhaps for nature, high density, mixed use is not necessarily the best alternative. 

I’m not sure what form a move to resilience in urban design might favour.  Most probably it will – and should – vary from place to place.  Decentralisation will have a role to play.  Certainly smaller centres, within, on, or beyond the edges of large cities, with a full range of services and amenities and a high level of self sufficiency are likely to offer more resilience to communities than centralised, hierarchical and interdependent services stretched over the entire city.  

In some places, well constructed and spacious high rise apartments set in extensive green spaces might work.  In others, terraced housing, each dwelling with a small garden, interconnected by pathways and roads to nearby community and commercial centres will be appropriate.  Traditional suburbs, perhaps scaled down, will have their place, providing private and public spaces to nurture families and nature.  High density suburbs with extensive parks, green belts, and generous transport corridors are another option. 

Whatever the form, the risk of disasters in our cities being compounded by crowding and mean design calls for putting resilience into the urban design equation.  The possibility of marginal long-term savings in fuel consumption and vehicle emissions used to justify constricting our urban places (and lives) may otherwise come at too high a cost.

 

 

[1]               This is hardly an original observation.  It was argued for Wellington in 1981, for example, by A Cibirowski in “Urban design and physical planning tools to make cities safer in earthquake prone areas”, cited in McKay, B (2005) “Plan Tectonics", Planning Quarterly

 

---